Commit
•
e79b90d
1
Parent(s):
66cf156
End of training
Browse files- README.md +3 -3
- all_results.json +13 -0
- eval_results.json +8 -0
- runs/May21_22-06-10_804716d6a90a/events.out.tfevents.1716330796.804716d6a90a.958.11 +3 -0
- train_results.json +8 -0
- trainer_state.json +1098 -0
README.md
CHANGED
@@ -22,7 +22,7 @@ model-index:
|
|
22 |
metrics:
|
23 |
- name: Accuracy
|
24 |
type: accuracy
|
25 |
-
value: 0.
|
26 |
---
|
27 |
|
28 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
@@ -32,8 +32,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
32 |
|
33 |
This model is a fine-tuned version of [facebook/deit-base-distilled-patch16-224](https://huggingface.co/facebook/deit-base-distilled-patch16-224) on the imagefolder dataset.
|
34 |
It achieves the following results on the evaluation set:
|
35 |
-
- Loss: 0.
|
36 |
-
- Accuracy: 0.
|
37 |
|
38 |
## Model description
|
39 |
|
|
|
22 |
metrics:
|
23 |
- name: Accuracy
|
24 |
type: accuracy
|
25 |
+
value: 0.8732394366197183
|
26 |
---
|
27 |
|
28 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
32 |
|
33 |
This model is a fine-tuned version of [facebook/deit-base-distilled-patch16-224](https://huggingface.co/facebook/deit-base-distilled-patch16-224) on the imagefolder dataset.
|
34 |
It achieves the following results on the evaluation set:
|
35 |
+
- Loss: 0.3816
|
36 |
+
- Accuracy: 0.8732
|
37 |
|
38 |
## Model description
|
39 |
|
all_results.json
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 92.3076923076923,
|
3 |
+
"eval_accuracy": 0.8732394366197183,
|
4 |
+
"eval_loss": 0.38157418370246887,
|
5 |
+
"eval_runtime": 1.0366,
|
6 |
+
"eval_samples_per_second": 68.495,
|
7 |
+
"eval_steps_per_second": 2.894,
|
8 |
+
"total_flos": 2.8402872494292173e+18,
|
9 |
+
"train_loss": 0.23691350758075713,
|
10 |
+
"train_runtime": 1617.7362,
|
11 |
+
"train_samples_per_second": 24.54,
|
12 |
+
"train_steps_per_second": 0.185
|
13 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 92.3076923076923,
|
3 |
+
"eval_accuracy": 0.8732394366197183,
|
4 |
+
"eval_loss": 0.38157418370246887,
|
5 |
+
"eval_runtime": 1.0366,
|
6 |
+
"eval_samples_per_second": 68.495,
|
7 |
+
"eval_steps_per_second": 2.894
|
8 |
+
}
|
runs/May21_22-06-10_804716d6a90a/events.out.tfevents.1716330796.804716d6a90a.958.11
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:89071e6a81832959586123b3efefe2fb86b548ad497ec3ef7e0677718dc622cf
|
3 |
+
size 411
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 92.3076923076923,
|
3 |
+
"total_flos": 2.8402872494292173e+18,
|
4 |
+
"train_loss": 0.23691350758075713,
|
5 |
+
"train_runtime": 1617.7362,
|
6 |
+
"train_samples_per_second": 24.54,
|
7 |
+
"train_steps_per_second": 0.185
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,1098 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.8732394366197183,
|
3 |
+
"best_model_checkpoint": "deit-base-distilled-patch16-224-65-fold1/checkpoint-201",
|
4 |
+
"epoch": 92.3076923076923,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 300,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.9230769230769231,
|
13 |
+
"eval_accuracy": 0.49295774647887325,
|
14 |
+
"eval_loss": 0.7887781858444214,
|
15 |
+
"eval_runtime": 1.0049,
|
16 |
+
"eval_samples_per_second": 70.656,
|
17 |
+
"eval_steps_per_second": 2.985,
|
18 |
+
"step": 3
|
19 |
+
},
|
20 |
+
{
|
21 |
+
"epoch": 1.8461538461538463,
|
22 |
+
"eval_accuracy": 0.5070422535211268,
|
23 |
+
"eval_loss": 0.7159498333930969,
|
24 |
+
"eval_runtime": 0.9626,
|
25 |
+
"eval_samples_per_second": 73.757,
|
26 |
+
"eval_steps_per_second": 3.116,
|
27 |
+
"step": 6
|
28 |
+
},
|
29 |
+
{
|
30 |
+
"epoch": 2.769230769230769,
|
31 |
+
"eval_accuracy": 0.5070422535211268,
|
32 |
+
"eval_loss": 0.7091380953788757,
|
33 |
+
"eval_runtime": 0.9551,
|
34 |
+
"eval_samples_per_second": 74.336,
|
35 |
+
"eval_steps_per_second": 3.141,
|
36 |
+
"step": 9
|
37 |
+
},
|
38 |
+
{
|
39 |
+
"epoch": 3.076923076923077,
|
40 |
+
"grad_norm": 5.692840576171875,
|
41 |
+
"learning_rate": 1.6666666666666667e-05,
|
42 |
+
"loss": 0.7703,
|
43 |
+
"step": 10
|
44 |
+
},
|
45 |
+
{
|
46 |
+
"epoch": 4.0,
|
47 |
+
"eval_accuracy": 0.5352112676056338,
|
48 |
+
"eval_loss": 0.6907549500465393,
|
49 |
+
"eval_runtime": 0.9654,
|
50 |
+
"eval_samples_per_second": 73.543,
|
51 |
+
"eval_steps_per_second": 3.107,
|
52 |
+
"step": 13
|
53 |
+
},
|
54 |
+
{
|
55 |
+
"epoch": 4.923076923076923,
|
56 |
+
"eval_accuracy": 0.6197183098591549,
|
57 |
+
"eval_loss": 0.6527045965194702,
|
58 |
+
"eval_runtime": 0.978,
|
59 |
+
"eval_samples_per_second": 72.596,
|
60 |
+
"eval_steps_per_second": 3.067,
|
61 |
+
"step": 16
|
62 |
+
},
|
63 |
+
{
|
64 |
+
"epoch": 5.846153846153846,
|
65 |
+
"eval_accuracy": 0.7323943661971831,
|
66 |
+
"eval_loss": 0.6236115097999573,
|
67 |
+
"eval_runtime": 0.972,
|
68 |
+
"eval_samples_per_second": 73.044,
|
69 |
+
"eval_steps_per_second": 3.086,
|
70 |
+
"step": 19
|
71 |
+
},
|
72 |
+
{
|
73 |
+
"epoch": 6.153846153846154,
|
74 |
+
"grad_norm": 4.818113327026367,
|
75 |
+
"learning_rate": 3.3333333333333335e-05,
|
76 |
+
"loss": 0.6435,
|
77 |
+
"step": 20
|
78 |
+
},
|
79 |
+
{
|
80 |
+
"epoch": 6.769230769230769,
|
81 |
+
"eval_accuracy": 0.6901408450704225,
|
82 |
+
"eval_loss": 0.6357130408287048,
|
83 |
+
"eval_runtime": 0.9775,
|
84 |
+
"eval_samples_per_second": 72.631,
|
85 |
+
"eval_steps_per_second": 3.069,
|
86 |
+
"step": 22
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 8.0,
|
90 |
+
"eval_accuracy": 0.704225352112676,
|
91 |
+
"eval_loss": 0.5442065596580505,
|
92 |
+
"eval_runtime": 1.0045,
|
93 |
+
"eval_samples_per_second": 70.683,
|
94 |
+
"eval_steps_per_second": 2.987,
|
95 |
+
"step": 26
|
96 |
+
},
|
97 |
+
{
|
98 |
+
"epoch": 8.923076923076923,
|
99 |
+
"eval_accuracy": 0.7183098591549296,
|
100 |
+
"eval_loss": 0.5448713898658752,
|
101 |
+
"eval_runtime": 0.9838,
|
102 |
+
"eval_samples_per_second": 72.166,
|
103 |
+
"eval_steps_per_second": 3.049,
|
104 |
+
"step": 29
|
105 |
+
},
|
106 |
+
{
|
107 |
+
"epoch": 9.23076923076923,
|
108 |
+
"grad_norm": 3.4911949634552,
|
109 |
+
"learning_rate": 5e-05,
|
110 |
+
"loss": 0.5366,
|
111 |
+
"step": 30
|
112 |
+
},
|
113 |
+
{
|
114 |
+
"epoch": 9.846153846153847,
|
115 |
+
"eval_accuracy": 0.7464788732394366,
|
116 |
+
"eval_loss": 0.5123764872550964,
|
117 |
+
"eval_runtime": 0.9995,
|
118 |
+
"eval_samples_per_second": 71.034,
|
119 |
+
"eval_steps_per_second": 3.001,
|
120 |
+
"step": 32
|
121 |
+
},
|
122 |
+
{
|
123 |
+
"epoch": 10.76923076923077,
|
124 |
+
"eval_accuracy": 0.704225352112676,
|
125 |
+
"eval_loss": 0.5028589367866516,
|
126 |
+
"eval_runtime": 0.9867,
|
127 |
+
"eval_samples_per_second": 71.958,
|
128 |
+
"eval_steps_per_second": 3.04,
|
129 |
+
"step": 35
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 12.0,
|
133 |
+
"eval_accuracy": 0.7183098591549296,
|
134 |
+
"eval_loss": 0.5486107468605042,
|
135 |
+
"eval_runtime": 0.9813,
|
136 |
+
"eval_samples_per_second": 72.355,
|
137 |
+
"eval_steps_per_second": 3.057,
|
138 |
+
"step": 39
|
139 |
+
},
|
140 |
+
{
|
141 |
+
"epoch": 12.307692307692308,
|
142 |
+
"grad_norm": 6.1214070320129395,
|
143 |
+
"learning_rate": 4.814814814814815e-05,
|
144 |
+
"loss": 0.4577,
|
145 |
+
"step": 40
|
146 |
+
},
|
147 |
+
{
|
148 |
+
"epoch": 12.923076923076923,
|
149 |
+
"eval_accuracy": 0.676056338028169,
|
150 |
+
"eval_loss": 0.5393596291542053,
|
151 |
+
"eval_runtime": 0.9861,
|
152 |
+
"eval_samples_per_second": 71.997,
|
153 |
+
"eval_steps_per_second": 3.042,
|
154 |
+
"step": 42
|
155 |
+
},
|
156 |
+
{
|
157 |
+
"epoch": 13.846153846153847,
|
158 |
+
"eval_accuracy": 0.7464788732394366,
|
159 |
+
"eval_loss": 0.5510964393615723,
|
160 |
+
"eval_runtime": 0.9874,
|
161 |
+
"eval_samples_per_second": 71.909,
|
162 |
+
"eval_steps_per_second": 3.038,
|
163 |
+
"step": 45
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 14.76923076923077,
|
167 |
+
"eval_accuracy": 0.6901408450704225,
|
168 |
+
"eval_loss": 0.5794062614440918,
|
169 |
+
"eval_runtime": 0.9882,
|
170 |
+
"eval_samples_per_second": 71.849,
|
171 |
+
"eval_steps_per_second": 3.036,
|
172 |
+
"step": 48
|
173 |
+
},
|
174 |
+
{
|
175 |
+
"epoch": 15.384615384615385,
|
176 |
+
"grad_norm": 4.4983015060424805,
|
177 |
+
"learning_rate": 4.62962962962963e-05,
|
178 |
+
"loss": 0.4187,
|
179 |
+
"step": 50
|
180 |
+
},
|
181 |
+
{
|
182 |
+
"epoch": 16.0,
|
183 |
+
"eval_accuracy": 0.7323943661971831,
|
184 |
+
"eval_loss": 0.53675776720047,
|
185 |
+
"eval_runtime": 1.0161,
|
186 |
+
"eval_samples_per_second": 69.878,
|
187 |
+
"eval_steps_per_second": 2.953,
|
188 |
+
"step": 52
|
189 |
+
},
|
190 |
+
{
|
191 |
+
"epoch": 16.923076923076923,
|
192 |
+
"eval_accuracy": 0.7887323943661971,
|
193 |
+
"eval_loss": 0.4677673578262329,
|
194 |
+
"eval_runtime": 0.9871,
|
195 |
+
"eval_samples_per_second": 71.928,
|
196 |
+
"eval_steps_per_second": 3.039,
|
197 |
+
"step": 55
|
198 |
+
},
|
199 |
+
{
|
200 |
+
"epoch": 17.846153846153847,
|
201 |
+
"eval_accuracy": 0.704225352112676,
|
202 |
+
"eval_loss": 0.6597200036048889,
|
203 |
+
"eval_runtime": 1.0221,
|
204 |
+
"eval_samples_per_second": 69.468,
|
205 |
+
"eval_steps_per_second": 2.935,
|
206 |
+
"step": 58
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 18.46153846153846,
|
210 |
+
"grad_norm": 4.643094539642334,
|
211 |
+
"learning_rate": 4.4444444444444447e-05,
|
212 |
+
"loss": 0.3542,
|
213 |
+
"step": 60
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 18.76923076923077,
|
217 |
+
"eval_accuracy": 0.8169014084507042,
|
218 |
+
"eval_loss": 0.4969327747821808,
|
219 |
+
"eval_runtime": 0.9908,
|
220 |
+
"eval_samples_per_second": 71.661,
|
221 |
+
"eval_steps_per_second": 3.028,
|
222 |
+
"step": 61
|
223 |
+
},
|
224 |
+
{
|
225 |
+
"epoch": 20.0,
|
226 |
+
"eval_accuracy": 0.7323943661971831,
|
227 |
+
"eval_loss": 0.7102980017662048,
|
228 |
+
"eval_runtime": 0.9798,
|
229 |
+
"eval_samples_per_second": 72.463,
|
230 |
+
"eval_steps_per_second": 3.062,
|
231 |
+
"step": 65
|
232 |
+
},
|
233 |
+
{
|
234 |
+
"epoch": 20.923076923076923,
|
235 |
+
"eval_accuracy": 0.7605633802816901,
|
236 |
+
"eval_loss": 0.49788960814476013,
|
237 |
+
"eval_runtime": 0.9857,
|
238 |
+
"eval_samples_per_second": 72.032,
|
239 |
+
"eval_steps_per_second": 3.044,
|
240 |
+
"step": 68
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 21.53846153846154,
|
244 |
+
"grad_norm": 3.1563220024108887,
|
245 |
+
"learning_rate": 4.259259259259259e-05,
|
246 |
+
"loss": 0.3057,
|
247 |
+
"step": 70
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 21.846153846153847,
|
251 |
+
"eval_accuracy": 0.7323943661971831,
|
252 |
+
"eval_loss": 0.5270604491233826,
|
253 |
+
"eval_runtime": 0.9884,
|
254 |
+
"eval_samples_per_second": 71.835,
|
255 |
+
"eval_steps_per_second": 3.035,
|
256 |
+
"step": 71
|
257 |
+
},
|
258 |
+
{
|
259 |
+
"epoch": 22.76923076923077,
|
260 |
+
"eval_accuracy": 0.7746478873239436,
|
261 |
+
"eval_loss": 0.5357353091239929,
|
262 |
+
"eval_runtime": 0.9814,
|
263 |
+
"eval_samples_per_second": 72.342,
|
264 |
+
"eval_steps_per_second": 3.057,
|
265 |
+
"step": 74
|
266 |
+
},
|
267 |
+
{
|
268 |
+
"epoch": 24.0,
|
269 |
+
"eval_accuracy": 0.7887323943661971,
|
270 |
+
"eval_loss": 0.48468706011772156,
|
271 |
+
"eval_runtime": 0.9993,
|
272 |
+
"eval_samples_per_second": 71.047,
|
273 |
+
"eval_steps_per_second": 3.002,
|
274 |
+
"step": 78
|
275 |
+
},
|
276 |
+
{
|
277 |
+
"epoch": 24.615384615384617,
|
278 |
+
"grad_norm": 6.270655632019043,
|
279 |
+
"learning_rate": 4.074074074074074e-05,
|
280 |
+
"loss": 0.2816,
|
281 |
+
"step": 80
|
282 |
+
},
|
283 |
+
{
|
284 |
+
"epoch": 24.923076923076923,
|
285 |
+
"eval_accuracy": 0.8309859154929577,
|
286 |
+
"eval_loss": 0.5424976944923401,
|
287 |
+
"eval_runtime": 0.9807,
|
288 |
+
"eval_samples_per_second": 72.4,
|
289 |
+
"eval_steps_per_second": 3.059,
|
290 |
+
"step": 81
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 25.846153846153847,
|
294 |
+
"eval_accuracy": 0.8028169014084507,
|
295 |
+
"eval_loss": 0.5239164233207703,
|
296 |
+
"eval_runtime": 1.0158,
|
297 |
+
"eval_samples_per_second": 69.897,
|
298 |
+
"eval_steps_per_second": 2.953,
|
299 |
+
"step": 84
|
300 |
+
},
|
301 |
+
{
|
302 |
+
"epoch": 26.76923076923077,
|
303 |
+
"eval_accuracy": 0.8309859154929577,
|
304 |
+
"eval_loss": 0.4140526354312897,
|
305 |
+
"eval_runtime": 0.9875,
|
306 |
+
"eval_samples_per_second": 71.897,
|
307 |
+
"eval_steps_per_second": 3.038,
|
308 |
+
"step": 87
|
309 |
+
},
|
310 |
+
{
|
311 |
+
"epoch": 27.692307692307693,
|
312 |
+
"grad_norm": 3.314903497695923,
|
313 |
+
"learning_rate": 3.888888888888889e-05,
|
314 |
+
"loss": 0.2881,
|
315 |
+
"step": 90
|
316 |
+
},
|
317 |
+
{
|
318 |
+
"epoch": 28.0,
|
319 |
+
"eval_accuracy": 0.8028169014084507,
|
320 |
+
"eval_loss": 0.4996558725833893,
|
321 |
+
"eval_runtime": 0.9975,
|
322 |
+
"eval_samples_per_second": 71.178,
|
323 |
+
"eval_steps_per_second": 3.008,
|
324 |
+
"step": 91
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 28.923076923076923,
|
328 |
+
"eval_accuracy": 0.8028169014084507,
|
329 |
+
"eval_loss": 0.42159879207611084,
|
330 |
+
"eval_runtime": 0.983,
|
331 |
+
"eval_samples_per_second": 72.227,
|
332 |
+
"eval_steps_per_second": 3.052,
|
333 |
+
"step": 94
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 29.846153846153847,
|
337 |
+
"eval_accuracy": 0.7887323943661971,
|
338 |
+
"eval_loss": 0.46680712699890137,
|
339 |
+
"eval_runtime": 0.9908,
|
340 |
+
"eval_samples_per_second": 71.663,
|
341 |
+
"eval_steps_per_second": 3.028,
|
342 |
+
"step": 97
|
343 |
+
},
|
344 |
+
{
|
345 |
+
"epoch": 30.76923076923077,
|
346 |
+
"grad_norm": 2.419731616973877,
|
347 |
+
"learning_rate": 3.7037037037037037e-05,
|
348 |
+
"loss": 0.2421,
|
349 |
+
"step": 100
|
350 |
+
},
|
351 |
+
{
|
352 |
+
"epoch": 30.76923076923077,
|
353 |
+
"eval_accuracy": 0.7887323943661971,
|
354 |
+
"eval_loss": 0.5904483199119568,
|
355 |
+
"eval_runtime": 0.9786,
|
356 |
+
"eval_samples_per_second": 72.553,
|
357 |
+
"eval_steps_per_second": 3.066,
|
358 |
+
"step": 100
|
359 |
+
},
|
360 |
+
{
|
361 |
+
"epoch": 32.0,
|
362 |
+
"eval_accuracy": 0.7746478873239436,
|
363 |
+
"eval_loss": 0.5239621996879578,
|
364 |
+
"eval_runtime": 0.9902,
|
365 |
+
"eval_samples_per_second": 71.7,
|
366 |
+
"eval_steps_per_second": 3.03,
|
367 |
+
"step": 104
|
368 |
+
},
|
369 |
+
{
|
370 |
+
"epoch": 32.92307692307692,
|
371 |
+
"eval_accuracy": 0.7605633802816901,
|
372 |
+
"eval_loss": 0.99370938539505,
|
373 |
+
"eval_runtime": 0.9918,
|
374 |
+
"eval_samples_per_second": 71.586,
|
375 |
+
"eval_steps_per_second": 3.025,
|
376 |
+
"step": 107
|
377 |
+
},
|
378 |
+
{
|
379 |
+
"epoch": 33.84615384615385,
|
380 |
+
"grad_norm": 2.3661601543426514,
|
381 |
+
"learning_rate": 3.518518518518519e-05,
|
382 |
+
"loss": 0.2402,
|
383 |
+
"step": 110
|
384 |
+
},
|
385 |
+
{
|
386 |
+
"epoch": 33.84615384615385,
|
387 |
+
"eval_accuracy": 0.8028169014084507,
|
388 |
+
"eval_loss": 0.4988890588283539,
|
389 |
+
"eval_runtime": 0.9897,
|
390 |
+
"eval_samples_per_second": 71.739,
|
391 |
+
"eval_steps_per_second": 3.031,
|
392 |
+
"step": 110
|
393 |
+
},
|
394 |
+
{
|
395 |
+
"epoch": 34.76923076923077,
|
396 |
+
"eval_accuracy": 0.7887323943661971,
|
397 |
+
"eval_loss": 0.7231740951538086,
|
398 |
+
"eval_runtime": 0.9872,
|
399 |
+
"eval_samples_per_second": 71.923,
|
400 |
+
"eval_steps_per_second": 3.039,
|
401 |
+
"step": 113
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 36.0,
|
405 |
+
"eval_accuracy": 0.8450704225352113,
|
406 |
+
"eval_loss": 0.4815123677253723,
|
407 |
+
"eval_runtime": 0.9925,
|
408 |
+
"eval_samples_per_second": 71.533,
|
409 |
+
"eval_steps_per_second": 3.023,
|
410 |
+
"step": 117
|
411 |
+
},
|
412 |
+
{
|
413 |
+
"epoch": 36.92307692307692,
|
414 |
+
"grad_norm": 2.6437292098999023,
|
415 |
+
"learning_rate": 3.3333333333333335e-05,
|
416 |
+
"loss": 0.1862,
|
417 |
+
"step": 120
|
418 |
+
},
|
419 |
+
{
|
420 |
+
"epoch": 36.92307692307692,
|
421 |
+
"eval_accuracy": 0.7746478873239436,
|
422 |
+
"eval_loss": 0.7431491017341614,
|
423 |
+
"eval_runtime": 0.9913,
|
424 |
+
"eval_samples_per_second": 71.622,
|
425 |
+
"eval_steps_per_second": 3.026,
|
426 |
+
"step": 120
|
427 |
+
},
|
428 |
+
{
|
429 |
+
"epoch": 37.84615384615385,
|
430 |
+
"eval_accuracy": 0.8028169014084507,
|
431 |
+
"eval_loss": 0.44340020418167114,
|
432 |
+
"eval_runtime": 0.9934,
|
433 |
+
"eval_samples_per_second": 71.475,
|
434 |
+
"eval_steps_per_second": 3.02,
|
435 |
+
"step": 123
|
436 |
+
},
|
437 |
+
{
|
438 |
+
"epoch": 38.76923076923077,
|
439 |
+
"eval_accuracy": 0.7887323943661971,
|
440 |
+
"eval_loss": 0.4759599268436432,
|
441 |
+
"eval_runtime": 0.9853,
|
442 |
+
"eval_samples_per_second": 72.061,
|
443 |
+
"eval_steps_per_second": 3.045,
|
444 |
+
"step": 126
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 40.0,
|
448 |
+
"grad_norm": 2.1550378799438477,
|
449 |
+
"learning_rate": 3.148148148148148e-05,
|
450 |
+
"loss": 0.1783,
|
451 |
+
"step": 130
|
452 |
+
},
|
453 |
+
{
|
454 |
+
"epoch": 40.0,
|
455 |
+
"eval_accuracy": 0.7887323943661971,
|
456 |
+
"eval_loss": 0.5006221532821655,
|
457 |
+
"eval_runtime": 0.9897,
|
458 |
+
"eval_samples_per_second": 71.737,
|
459 |
+
"eval_steps_per_second": 3.031,
|
460 |
+
"step": 130
|
461 |
+
},
|
462 |
+
{
|
463 |
+
"epoch": 40.92307692307692,
|
464 |
+
"eval_accuracy": 0.7887323943661971,
|
465 |
+
"eval_loss": 0.49856680631637573,
|
466 |
+
"eval_runtime": 0.9788,
|
467 |
+
"eval_samples_per_second": 72.535,
|
468 |
+
"eval_steps_per_second": 3.065,
|
469 |
+
"step": 133
|
470 |
+
},
|
471 |
+
{
|
472 |
+
"epoch": 41.84615384615385,
|
473 |
+
"eval_accuracy": 0.7887323943661971,
|
474 |
+
"eval_loss": 0.794731616973877,
|
475 |
+
"eval_runtime": 0.9805,
|
476 |
+
"eval_samples_per_second": 72.414,
|
477 |
+
"eval_steps_per_second": 3.06,
|
478 |
+
"step": 136
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 42.76923076923077,
|
482 |
+
"eval_accuracy": 0.8309859154929577,
|
483 |
+
"eval_loss": 0.48966798186302185,
|
484 |
+
"eval_runtime": 0.9856,
|
485 |
+
"eval_samples_per_second": 72.035,
|
486 |
+
"eval_steps_per_second": 3.044,
|
487 |
+
"step": 139
|
488 |
+
},
|
489 |
+
{
|
490 |
+
"epoch": 43.07692307692308,
|
491 |
+
"grad_norm": 4.09879207611084,
|
492 |
+
"learning_rate": 2.962962962962963e-05,
|
493 |
+
"loss": 0.1685,
|
494 |
+
"step": 140
|
495 |
+
},
|
496 |
+
{
|
497 |
+
"epoch": 44.0,
|
498 |
+
"eval_accuracy": 0.7605633802816901,
|
499 |
+
"eval_loss": 0.7500377893447876,
|
500 |
+
"eval_runtime": 0.9859,
|
501 |
+
"eval_samples_per_second": 72.013,
|
502 |
+
"eval_steps_per_second": 3.043,
|
503 |
+
"step": 143
|
504 |
+
},
|
505 |
+
{
|
506 |
+
"epoch": 44.92307692307692,
|
507 |
+
"eval_accuracy": 0.7887323943661971,
|
508 |
+
"eval_loss": 0.6052554249763489,
|
509 |
+
"eval_runtime": 0.9867,
|
510 |
+
"eval_samples_per_second": 71.954,
|
511 |
+
"eval_steps_per_second": 3.04,
|
512 |
+
"step": 146
|
513 |
+
},
|
514 |
+
{
|
515 |
+
"epoch": 45.84615384615385,
|
516 |
+
"eval_accuracy": 0.8169014084507042,
|
517 |
+
"eval_loss": 0.4776698350906372,
|
518 |
+
"eval_runtime": 0.9829,
|
519 |
+
"eval_samples_per_second": 72.233,
|
520 |
+
"eval_steps_per_second": 3.052,
|
521 |
+
"step": 149
|
522 |
+
},
|
523 |
+
{
|
524 |
+
"epoch": 46.15384615384615,
|
525 |
+
"grad_norm": 2.7736315727233887,
|
526 |
+
"learning_rate": 2.777777777777778e-05,
|
527 |
+
"loss": 0.1779,
|
528 |
+
"step": 150
|
529 |
+
},
|
530 |
+
{
|
531 |
+
"epoch": 46.76923076923077,
|
532 |
+
"eval_accuracy": 0.7746478873239436,
|
533 |
+
"eval_loss": 0.5799605250358582,
|
534 |
+
"eval_runtime": 0.9783,
|
535 |
+
"eval_samples_per_second": 72.578,
|
536 |
+
"eval_steps_per_second": 3.067,
|
537 |
+
"step": 152
|
538 |
+
},
|
539 |
+
{
|
540 |
+
"epoch": 48.0,
|
541 |
+
"eval_accuracy": 0.8450704225352113,
|
542 |
+
"eval_loss": 0.4681435227394104,
|
543 |
+
"eval_runtime": 0.984,
|
544 |
+
"eval_samples_per_second": 72.155,
|
545 |
+
"eval_steps_per_second": 3.049,
|
546 |
+
"step": 156
|
547 |
+
},
|
548 |
+
{
|
549 |
+
"epoch": 48.92307692307692,
|
550 |
+
"eval_accuracy": 0.8028169014084507,
|
551 |
+
"eval_loss": 0.7729381322860718,
|
552 |
+
"eval_runtime": 0.9844,
|
553 |
+
"eval_samples_per_second": 72.123,
|
554 |
+
"eval_steps_per_second": 3.047,
|
555 |
+
"step": 159
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 49.23076923076923,
|
559 |
+
"grad_norm": 6.783754825592041,
|
560 |
+
"learning_rate": 2.5925925925925925e-05,
|
561 |
+
"loss": 0.1502,
|
562 |
+
"step": 160
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 49.84615384615385,
|
566 |
+
"eval_accuracy": 0.8028169014084507,
|
567 |
+
"eval_loss": 0.6486974358558655,
|
568 |
+
"eval_runtime": 0.9835,
|
569 |
+
"eval_samples_per_second": 72.191,
|
570 |
+
"eval_steps_per_second": 3.05,
|
571 |
+
"step": 162
|
572 |
+
},
|
573 |
+
{
|
574 |
+
"epoch": 50.76923076923077,
|
575 |
+
"eval_accuracy": 0.8169014084507042,
|
576 |
+
"eval_loss": 0.5224327445030212,
|
577 |
+
"eval_runtime": 0.982,
|
578 |
+
"eval_samples_per_second": 72.3,
|
579 |
+
"eval_steps_per_second": 3.055,
|
580 |
+
"step": 165
|
581 |
+
},
|
582 |
+
{
|
583 |
+
"epoch": 52.0,
|
584 |
+
"eval_accuracy": 0.8028169014084507,
|
585 |
+
"eval_loss": 0.7017256021499634,
|
586 |
+
"eval_runtime": 1.0037,
|
587 |
+
"eval_samples_per_second": 70.739,
|
588 |
+
"eval_steps_per_second": 2.989,
|
589 |
+
"step": 169
|
590 |
+
},
|
591 |
+
{
|
592 |
+
"epoch": 52.30769230769231,
|
593 |
+
"grad_norm": 2.59287428855896,
|
594 |
+
"learning_rate": 2.4074074074074074e-05,
|
595 |
+
"loss": 0.1586,
|
596 |
+
"step": 170
|
597 |
+
},
|
598 |
+
{
|
599 |
+
"epoch": 52.92307692307692,
|
600 |
+
"eval_accuracy": 0.8028169014084507,
|
601 |
+
"eval_loss": 0.603446364402771,
|
602 |
+
"eval_runtime": 1.0504,
|
603 |
+
"eval_samples_per_second": 67.596,
|
604 |
+
"eval_steps_per_second": 2.856,
|
605 |
+
"step": 172
|
606 |
+
},
|
607 |
+
{
|
608 |
+
"epoch": 53.84615384615385,
|
609 |
+
"eval_accuracy": 0.8028169014084507,
|
610 |
+
"eval_loss": 0.5791419744491577,
|
611 |
+
"eval_runtime": 1.0306,
|
612 |
+
"eval_samples_per_second": 68.891,
|
613 |
+
"eval_steps_per_second": 2.911,
|
614 |
+
"step": 175
|
615 |
+
},
|
616 |
+
{
|
617 |
+
"epoch": 54.76923076923077,
|
618 |
+
"eval_accuracy": 0.8169014084507042,
|
619 |
+
"eval_loss": 0.565058708190918,
|
620 |
+
"eval_runtime": 0.9801,
|
621 |
+
"eval_samples_per_second": 72.445,
|
622 |
+
"eval_steps_per_second": 3.061,
|
623 |
+
"step": 178
|
624 |
+
},
|
625 |
+
{
|
626 |
+
"epoch": 55.38461538461539,
|
627 |
+
"grad_norm": 3.497166872024536,
|
628 |
+
"learning_rate": 2.2222222222222223e-05,
|
629 |
+
"loss": 0.134,
|
630 |
+
"step": 180
|
631 |
+
},
|
632 |
+
{
|
633 |
+
"epoch": 56.0,
|
634 |
+
"eval_accuracy": 0.8028169014084507,
|
635 |
+
"eval_loss": 0.486237496137619,
|
636 |
+
"eval_runtime": 0.98,
|
637 |
+
"eval_samples_per_second": 72.446,
|
638 |
+
"eval_steps_per_second": 3.061,
|
639 |
+
"step": 182
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 56.92307692307692,
|
643 |
+
"eval_accuracy": 0.8169014084507042,
|
644 |
+
"eval_loss": 0.6750636100769043,
|
645 |
+
"eval_runtime": 0.9858,
|
646 |
+
"eval_samples_per_second": 72.026,
|
647 |
+
"eval_steps_per_second": 3.043,
|
648 |
+
"step": 185
|
649 |
+
},
|
650 |
+
{
|
651 |
+
"epoch": 57.84615384615385,
|
652 |
+
"eval_accuracy": 0.8169014084507042,
|
653 |
+
"eval_loss": 0.5924672484397888,
|
654 |
+
"eval_runtime": 1.0279,
|
655 |
+
"eval_samples_per_second": 69.075,
|
656 |
+
"eval_steps_per_second": 2.919,
|
657 |
+
"step": 188
|
658 |
+
},
|
659 |
+
{
|
660 |
+
"epoch": 58.46153846153846,
|
661 |
+
"grad_norm": 1.8251292705535889,
|
662 |
+
"learning_rate": 2.037037037037037e-05,
|
663 |
+
"loss": 0.1602,
|
664 |
+
"step": 190
|
665 |
+
},
|
666 |
+
{
|
667 |
+
"epoch": 58.76923076923077,
|
668 |
+
"eval_accuracy": 0.8450704225352113,
|
669 |
+
"eval_loss": 0.39819350838661194,
|
670 |
+
"eval_runtime": 0.9991,
|
671 |
+
"eval_samples_per_second": 71.061,
|
672 |
+
"eval_steps_per_second": 3.003,
|
673 |
+
"step": 191
|
674 |
+
},
|
675 |
+
{
|
676 |
+
"epoch": 60.0,
|
677 |
+
"eval_accuracy": 0.7887323943661971,
|
678 |
+
"eval_loss": 0.5968966484069824,
|
679 |
+
"eval_runtime": 1.0038,
|
680 |
+
"eval_samples_per_second": 70.73,
|
681 |
+
"eval_steps_per_second": 2.989,
|
682 |
+
"step": 195
|
683 |
+
},
|
684 |
+
{
|
685 |
+
"epoch": 60.92307692307692,
|
686 |
+
"eval_accuracy": 0.7887323943661971,
|
687 |
+
"eval_loss": 0.5720605254173279,
|
688 |
+
"eval_runtime": 0.9693,
|
689 |
+
"eval_samples_per_second": 73.252,
|
690 |
+
"eval_steps_per_second": 3.095,
|
691 |
+
"step": 198
|
692 |
+
},
|
693 |
+
{
|
694 |
+
"epoch": 61.53846153846154,
|
695 |
+
"grad_norm": 1.8045357465744019,
|
696 |
+
"learning_rate": 1.8518518518518518e-05,
|
697 |
+
"loss": 0.1217,
|
698 |
+
"step": 200
|
699 |
+
},
|
700 |
+
{
|
701 |
+
"epoch": 61.84615384615385,
|
702 |
+
"eval_accuracy": 0.8732394366197183,
|
703 |
+
"eval_loss": 0.38157418370246887,
|
704 |
+
"eval_runtime": 0.9841,
|
705 |
+
"eval_samples_per_second": 72.148,
|
706 |
+
"eval_steps_per_second": 3.048,
|
707 |
+
"step": 201
|
708 |
+
},
|
709 |
+
{
|
710 |
+
"epoch": 62.76923076923077,
|
711 |
+
"eval_accuracy": 0.8309859154929577,
|
712 |
+
"eval_loss": 0.4109618365764618,
|
713 |
+
"eval_runtime": 0.9741,
|
714 |
+
"eval_samples_per_second": 72.89,
|
715 |
+
"eval_steps_per_second": 3.08,
|
716 |
+
"step": 204
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 64.0,
|
720 |
+
"eval_accuracy": 0.7887323943661971,
|
721 |
+
"eval_loss": 0.6715837717056274,
|
722 |
+
"eval_runtime": 0.9934,
|
723 |
+
"eval_samples_per_second": 71.469,
|
724 |
+
"eval_steps_per_second": 3.02,
|
725 |
+
"step": 208
|
726 |
+
},
|
727 |
+
{
|
728 |
+
"epoch": 64.61538461538461,
|
729 |
+
"grad_norm": 4.87626838684082,
|
730 |
+
"learning_rate": 1.6666666666666667e-05,
|
731 |
+
"loss": 0.1274,
|
732 |
+
"step": 210
|
733 |
+
},
|
734 |
+
{
|
735 |
+
"epoch": 64.92307692307692,
|
736 |
+
"eval_accuracy": 0.8732394366197183,
|
737 |
+
"eval_loss": 0.34985408186912537,
|
738 |
+
"eval_runtime": 0.9766,
|
739 |
+
"eval_samples_per_second": 72.704,
|
740 |
+
"eval_steps_per_second": 3.072,
|
741 |
+
"step": 211
|
742 |
+
},
|
743 |
+
{
|
744 |
+
"epoch": 65.84615384615384,
|
745 |
+
"eval_accuracy": 0.8169014084507042,
|
746 |
+
"eval_loss": 0.36712145805358887,
|
747 |
+
"eval_runtime": 0.9825,
|
748 |
+
"eval_samples_per_second": 72.266,
|
749 |
+
"eval_steps_per_second": 3.053,
|
750 |
+
"step": 214
|
751 |
+
},
|
752 |
+
{
|
753 |
+
"epoch": 66.76923076923077,
|
754 |
+
"eval_accuracy": 0.7887323943661971,
|
755 |
+
"eval_loss": 0.5317980051040649,
|
756 |
+
"eval_runtime": 0.9694,
|
757 |
+
"eval_samples_per_second": 73.245,
|
758 |
+
"eval_steps_per_second": 3.095,
|
759 |
+
"step": 217
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 67.6923076923077,
|
763 |
+
"grad_norm": 1.64058518409729,
|
764 |
+
"learning_rate": 1.4814814814814815e-05,
|
765 |
+
"loss": 0.1277,
|
766 |
+
"step": 220
|
767 |
+
},
|
768 |
+
{
|
769 |
+
"epoch": 68.0,
|
770 |
+
"eval_accuracy": 0.7887323943661971,
|
771 |
+
"eval_loss": 0.6733711361885071,
|
772 |
+
"eval_runtime": 0.9874,
|
773 |
+
"eval_samples_per_second": 71.905,
|
774 |
+
"eval_steps_per_second": 3.038,
|
775 |
+
"step": 221
|
776 |
+
},
|
777 |
+
{
|
778 |
+
"epoch": 68.92307692307692,
|
779 |
+
"eval_accuracy": 0.8028169014084507,
|
780 |
+
"eval_loss": 0.47263458371162415,
|
781 |
+
"eval_runtime": 0.9771,
|
782 |
+
"eval_samples_per_second": 72.665,
|
783 |
+
"eval_steps_per_second": 3.07,
|
784 |
+
"step": 224
|
785 |
+
},
|
786 |
+
{
|
787 |
+
"epoch": 69.84615384615384,
|
788 |
+
"eval_accuracy": 0.8169014084507042,
|
789 |
+
"eval_loss": 0.4311037063598633,
|
790 |
+
"eval_runtime": 0.9776,
|
791 |
+
"eval_samples_per_second": 72.63,
|
792 |
+
"eval_steps_per_second": 3.069,
|
793 |
+
"step": 227
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 70.76923076923077,
|
797 |
+
"grad_norm": 2.334254026412964,
|
798 |
+
"learning_rate": 1.2962962962962962e-05,
|
799 |
+
"loss": 0.1232,
|
800 |
+
"step": 230
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 70.76923076923077,
|
804 |
+
"eval_accuracy": 0.7746478873239436,
|
805 |
+
"eval_loss": 0.7072361707687378,
|
806 |
+
"eval_runtime": 0.969,
|
807 |
+
"eval_samples_per_second": 73.27,
|
808 |
+
"eval_steps_per_second": 3.096,
|
809 |
+
"step": 230
|
810 |
+
},
|
811 |
+
{
|
812 |
+
"epoch": 72.0,
|
813 |
+
"eval_accuracy": 0.7887323943661971,
|
814 |
+
"eval_loss": 0.5859442949295044,
|
815 |
+
"eval_runtime": 0.9757,
|
816 |
+
"eval_samples_per_second": 72.767,
|
817 |
+
"eval_steps_per_second": 3.075,
|
818 |
+
"step": 234
|
819 |
+
},
|
820 |
+
{
|
821 |
+
"epoch": 72.92307692307692,
|
822 |
+
"eval_accuracy": 0.8309859154929577,
|
823 |
+
"eval_loss": 0.3757818341255188,
|
824 |
+
"eval_runtime": 0.9776,
|
825 |
+
"eval_samples_per_second": 72.628,
|
826 |
+
"eval_steps_per_second": 3.069,
|
827 |
+
"step": 237
|
828 |
+
},
|
829 |
+
{
|
830 |
+
"epoch": 73.84615384615384,
|
831 |
+
"grad_norm": 1.5687333345413208,
|
832 |
+
"learning_rate": 1.1111111111111112e-05,
|
833 |
+
"loss": 0.1293,
|
834 |
+
"step": 240
|
835 |
+
},
|
836 |
+
{
|
837 |
+
"epoch": 73.84615384615384,
|
838 |
+
"eval_accuracy": 0.8450704225352113,
|
839 |
+
"eval_loss": 0.3672768175601959,
|
840 |
+
"eval_runtime": 0.977,
|
841 |
+
"eval_samples_per_second": 72.668,
|
842 |
+
"eval_steps_per_second": 3.07,
|
843 |
+
"step": 240
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 74.76923076923077,
|
847 |
+
"eval_accuracy": 0.8591549295774648,
|
848 |
+
"eval_loss": 0.367299884557724,
|
849 |
+
"eval_runtime": 0.9773,
|
850 |
+
"eval_samples_per_second": 72.646,
|
851 |
+
"eval_steps_per_second": 3.07,
|
852 |
+
"step": 243
|
853 |
+
},
|
854 |
+
{
|
855 |
+
"epoch": 76.0,
|
856 |
+
"eval_accuracy": 0.8169014084507042,
|
857 |
+
"eval_loss": 0.47515830397605896,
|
858 |
+
"eval_runtime": 0.9705,
|
859 |
+
"eval_samples_per_second": 73.158,
|
860 |
+
"eval_steps_per_second": 3.091,
|
861 |
+
"step": 247
|
862 |
+
},
|
863 |
+
{
|
864 |
+
"epoch": 76.92307692307692,
|
865 |
+
"grad_norm": 2.638277053833008,
|
866 |
+
"learning_rate": 9.259259259259259e-06,
|
867 |
+
"loss": 0.1117,
|
868 |
+
"step": 250
|
869 |
+
},
|
870 |
+
{
|
871 |
+
"epoch": 76.92307692307692,
|
872 |
+
"eval_accuracy": 0.8309859154929577,
|
873 |
+
"eval_loss": 0.44498294591903687,
|
874 |
+
"eval_runtime": 0.9801,
|
875 |
+
"eval_samples_per_second": 72.44,
|
876 |
+
"eval_steps_per_second": 3.061,
|
877 |
+
"step": 250
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 77.84615384615384,
|
881 |
+
"eval_accuracy": 0.8450704225352113,
|
882 |
+
"eval_loss": 0.44371137022972107,
|
883 |
+
"eval_runtime": 0.9778,
|
884 |
+
"eval_samples_per_second": 72.608,
|
885 |
+
"eval_steps_per_second": 3.068,
|
886 |
+
"step": 253
|
887 |
+
},
|
888 |
+
{
|
889 |
+
"epoch": 78.76923076923077,
|
890 |
+
"eval_accuracy": 0.8309859154929577,
|
891 |
+
"eval_loss": 0.4329843521118164,
|
892 |
+
"eval_runtime": 0.9979,
|
893 |
+
"eval_samples_per_second": 71.146,
|
894 |
+
"eval_steps_per_second": 3.006,
|
895 |
+
"step": 256
|
896 |
+
},
|
897 |
+
{
|
898 |
+
"epoch": 80.0,
|
899 |
+
"grad_norm": 2.258843421936035,
|
900 |
+
"learning_rate": 7.4074074074074075e-06,
|
901 |
+
"loss": 0.1092,
|
902 |
+
"step": 260
|
903 |
+
},
|
904 |
+
{
|
905 |
+
"epoch": 80.0,
|
906 |
+
"eval_accuracy": 0.8169014084507042,
|
907 |
+
"eval_loss": 0.5095254182815552,
|
908 |
+
"eval_runtime": 0.9679,
|
909 |
+
"eval_samples_per_second": 73.356,
|
910 |
+
"eval_steps_per_second": 3.1,
|
911 |
+
"step": 260
|
912 |
+
},
|
913 |
+
{
|
914 |
+
"epoch": 80.92307692307692,
|
915 |
+
"eval_accuracy": 0.8169014084507042,
|
916 |
+
"eval_loss": 0.49475494027137756,
|
917 |
+
"eval_runtime": 0.9923,
|
918 |
+
"eval_samples_per_second": 71.549,
|
919 |
+
"eval_steps_per_second": 3.023,
|
920 |
+
"step": 263
|
921 |
+
},
|
922 |
+
{
|
923 |
+
"epoch": 81.84615384615384,
|
924 |
+
"eval_accuracy": 0.8591549295774648,
|
925 |
+
"eval_loss": 0.413480281829834,
|
926 |
+
"eval_runtime": 0.9737,
|
927 |
+
"eval_samples_per_second": 72.914,
|
928 |
+
"eval_steps_per_second": 3.081,
|
929 |
+
"step": 266
|
930 |
+
},
|
931 |
+
{
|
932 |
+
"epoch": 82.76923076923077,
|
933 |
+
"eval_accuracy": 0.8450704225352113,
|
934 |
+
"eval_loss": 0.4189954698085785,
|
935 |
+
"eval_runtime": 0.9819,
|
936 |
+
"eval_samples_per_second": 72.311,
|
937 |
+
"eval_steps_per_second": 3.055,
|
938 |
+
"step": 269
|
939 |
+
},
|
940 |
+
{
|
941 |
+
"epoch": 83.07692307692308,
|
942 |
+
"grad_norm": 2.083444595336914,
|
943 |
+
"learning_rate": 5.555555555555556e-06,
|
944 |
+
"loss": 0.1151,
|
945 |
+
"step": 270
|
946 |
+
},
|
947 |
+
{
|
948 |
+
"epoch": 84.0,
|
949 |
+
"eval_accuracy": 0.8732394366197183,
|
950 |
+
"eval_loss": 0.41941961646080017,
|
951 |
+
"eval_runtime": 0.9662,
|
952 |
+
"eval_samples_per_second": 73.482,
|
953 |
+
"eval_steps_per_second": 3.105,
|
954 |
+
"step": 273
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 84.92307692307692,
|
958 |
+
"eval_accuracy": 0.8309859154929577,
|
959 |
+
"eval_loss": 0.43557149171829224,
|
960 |
+
"eval_runtime": 0.9875,
|
961 |
+
"eval_samples_per_second": 71.901,
|
962 |
+
"eval_steps_per_second": 3.038,
|
963 |
+
"step": 276
|
964 |
+
},
|
965 |
+
{
|
966 |
+
"epoch": 85.84615384615384,
|
967 |
+
"eval_accuracy": 0.8028169014084507,
|
968 |
+
"eval_loss": 0.46227431297302246,
|
969 |
+
"eval_runtime": 0.9652,
|
970 |
+
"eval_samples_per_second": 73.558,
|
971 |
+
"eval_steps_per_second": 3.108,
|
972 |
+
"step": 279
|
973 |
+
},
|
974 |
+
{
|
975 |
+
"epoch": 86.15384615384616,
|
976 |
+
"grad_norm": 3.552259683609009,
|
977 |
+
"learning_rate": 3.7037037037037037e-06,
|
978 |
+
"loss": 0.1085,
|
979 |
+
"step": 280
|
980 |
+
},
|
981 |
+
{
|
982 |
+
"epoch": 86.76923076923077,
|
983 |
+
"eval_accuracy": 0.8309859154929577,
|
984 |
+
"eval_loss": 0.48452523350715637,
|
985 |
+
"eval_runtime": 0.9766,
|
986 |
+
"eval_samples_per_second": 72.702,
|
987 |
+
"eval_steps_per_second": 3.072,
|
988 |
+
"step": 282
|
989 |
+
},
|
990 |
+
{
|
991 |
+
"epoch": 88.0,
|
992 |
+
"eval_accuracy": 0.8169014084507042,
|
993 |
+
"eval_loss": 0.49984505772590637,
|
994 |
+
"eval_runtime": 0.9733,
|
995 |
+
"eval_samples_per_second": 72.949,
|
996 |
+
"eval_steps_per_second": 3.082,
|
997 |
+
"step": 286
|
998 |
+
},
|
999 |
+
{
|
1000 |
+
"epoch": 88.92307692307692,
|
1001 |
+
"eval_accuracy": 0.8028169014084507,
|
1002 |
+
"eval_loss": 0.5181298851966858,
|
1003 |
+
"eval_runtime": 0.9778,
|
1004 |
+
"eval_samples_per_second": 72.612,
|
1005 |
+
"eval_steps_per_second": 3.068,
|
1006 |
+
"step": 289
|
1007 |
+
},
|
1008 |
+
{
|
1009 |
+
"epoch": 89.23076923076923,
|
1010 |
+
"grad_norm": 3.29520583152771,
|
1011 |
+
"learning_rate": 1.8518518518518519e-06,
|
1012 |
+
"loss": 0.0908,
|
1013 |
+
"step": 290
|
1014 |
+
},
|
1015 |
+
{
|
1016 |
+
"epoch": 89.84615384615384,
|
1017 |
+
"eval_accuracy": 0.8169014084507042,
|
1018 |
+
"eval_loss": 0.5373069643974304,
|
1019 |
+
"eval_runtime": 0.9643,
|
1020 |
+
"eval_samples_per_second": 73.627,
|
1021 |
+
"eval_steps_per_second": 3.111,
|
1022 |
+
"step": 292
|
1023 |
+
},
|
1024 |
+
{
|
1025 |
+
"epoch": 90.76923076923077,
|
1026 |
+
"eval_accuracy": 0.8169014084507042,
|
1027 |
+
"eval_loss": 0.5465375185012817,
|
1028 |
+
"eval_runtime": 0.9916,
|
1029 |
+
"eval_samples_per_second": 71.6,
|
1030 |
+
"eval_steps_per_second": 3.025,
|
1031 |
+
"step": 295
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 92.0,
|
1035 |
+
"eval_accuracy": 0.8169014084507042,
|
1036 |
+
"eval_loss": 0.5421547889709473,
|
1037 |
+
"eval_runtime": 0.9804,
|
1038 |
+
"eval_samples_per_second": 72.422,
|
1039 |
+
"eval_steps_per_second": 3.06,
|
1040 |
+
"step": 299
|
1041 |
+
},
|
1042 |
+
{
|
1043 |
+
"epoch": 92.3076923076923,
|
1044 |
+
"grad_norm": 1.8515039682388306,
|
1045 |
+
"learning_rate": 0.0,
|
1046 |
+
"loss": 0.0902,
|
1047 |
+
"step": 300
|
1048 |
+
},
|
1049 |
+
{
|
1050 |
+
"epoch": 92.3076923076923,
|
1051 |
+
"eval_accuracy": 0.8169014084507042,
|
1052 |
+
"eval_loss": 0.5417219996452332,
|
1053 |
+
"eval_runtime": 0.9779,
|
1054 |
+
"eval_samples_per_second": 72.602,
|
1055 |
+
"eval_steps_per_second": 3.068,
|
1056 |
+
"step": 300
|
1057 |
+
},
|
1058 |
+
{
|
1059 |
+
"epoch": 92.3076923076923,
|
1060 |
+
"step": 300,
|
1061 |
+
"total_flos": 2.8402872494292173e+18,
|
1062 |
+
"train_loss": 0.23691350758075713,
|
1063 |
+
"train_runtime": 1617.7362,
|
1064 |
+
"train_samples_per_second": 24.54,
|
1065 |
+
"train_steps_per_second": 0.185
|
1066 |
+
},
|
1067 |
+
{
|
1068 |
+
"epoch": 92.3076923076923,
|
1069 |
+
"eval_accuracy": 0.8732394366197183,
|
1070 |
+
"eval_loss": 0.38157418370246887,
|
1071 |
+
"eval_runtime": 1.0366,
|
1072 |
+
"eval_samples_per_second": 68.495,
|
1073 |
+
"eval_steps_per_second": 2.894,
|
1074 |
+
"step": 300
|
1075 |
+
}
|
1076 |
+
],
|
1077 |
+
"logging_steps": 10,
|
1078 |
+
"max_steps": 300,
|
1079 |
+
"num_input_tokens_seen": 0,
|
1080 |
+
"num_train_epochs": 100,
|
1081 |
+
"save_steps": 500,
|
1082 |
+
"stateful_callbacks": {
|
1083 |
+
"TrainerControl": {
|
1084 |
+
"args": {
|
1085 |
+
"should_epoch_stop": false,
|
1086 |
+
"should_evaluate": false,
|
1087 |
+
"should_log": false,
|
1088 |
+
"should_save": true,
|
1089 |
+
"should_training_stop": true
|
1090 |
+
},
|
1091 |
+
"attributes": {}
|
1092 |
+
}
|
1093 |
+
},
|
1094 |
+
"total_flos": 2.8402872494292173e+18,
|
1095 |
+
"train_batch_size": 32,
|
1096 |
+
"trial_name": null,
|
1097 |
+
"trial_params": null
|
1098 |
+
}
|